Journal of Natural Prodvcts
1316
VO~. S7, NO.9, pp. 1316-1319, St.ptember 1994
BINDING OF QUINOLIZIDINE ALKALOIDS TO NICOTINIC AND MUSCARINIC ACETYLCHOLINE RECEPTORS THORSTEN %HMELLER,
-TINA
SAUERWEIN, FRANK SPORER, MK3lAEL WINK,*
Uniwsitat Heidelhrg, lnstitutfur Pharmazerrtische Biologie, Im Netrenhkmer Feld364, 69120 Heidelhrg, Germany and WALTER E. MULLER
Abteilung Psychpbamkologie, Zentralinstitutfur Seelische Gaudheit, I 5, 681 67 Mannheim, G m y ABsmcr.-Fourteen quinolizidinealkaloids,isolated from Lupinus albus, L. mutabilis,and ANlgYris faetih, were analyzed for their affinity for nicotinic and/or muscarinic acetylcholine receptors.Ofthe compounds tested, the a-pyridones,N-methylcytisine and cytisine, showed the highest affinities at the nicotinic receptor, while several quinolizidine alkaloid types were especially active at the muscarinic receptor.
Quinolitidine (lupine) alkaloids are characteristic secondary metabolites in many taxaof the subfamily Papilionoideae (family Leguminosae)and over 150structures have been described (1,2). Quinolitidine alkaloids are important for the well-being of the plants producing them (2-5), since they protect against herbivores, such as insects and grazing mammals. Other functions involve interactions with bacteria, fungi, viruses, and competing plants (2-8). Various pharmacological and toxicological properties have been attributed to particular lupine alkaloids such as antipyretic, antiinflammatory, CNS depressant, anti-arrhythmic, respiratory depressant and stimulant, uterotonic, diuretic, hypoglycemic,hypotensive, hallucinogenic, antidiabetic and mutagenic properties (1). Quinolitidine alkaloids exhibit both general vertebrate and insect toxicity(1,2,6,7). Modulationofacetylcholine receptors and ion channels (Na', K') by quinolizidine alkaloids has been suggested as a mechanism and explanation for toxicity and some of their pharmacological properties (1-8). As most quinolizidine alkaloids are not commercially available, only a few like sparteine, cytisine, and lupinine have been investigated biochemically in any detail. Because plants contain a complex mixture ofalkaloids, it would be interesting to know how much each single alka-
loidal component contributes to the toxicity and pharmacology of the mixture and, futthermore, which molecular targets are affected. We have isolated or purchased fourteen quinolizidine alkaloids, which are the major constituents of Lupinus albus, L. mutabilis, L. luteus, L. angustifolius, Anagyrisfwtiak, and Laburnum anagyroiah. In this communication, we describe the interaction of these alkaloids with two acetylcholine receptors (ACh) known to be stimulated by cytisine and sparteine (12). The activity of quinolizidine alkaloids at the nicotinic and muscarinic AChreceptors was assayed by measuring the displacement ofradiolabeled ligands, 3Hnicotine and 3H-quinuclidinyl benzilate (3H-QNB), respectively, by employing a rapid filtration technique (9).All 14 pure quinolitidine alkaloids were assayed in concentrations between 0.5 nM and 10 mM. Typical displacement curves are illustrated in Figures 1A and 1B. From these graphs, IC,, values were calculated as the concentrations which displace 50% of the specifically bound ligands (Table 1). At the nicotinic receptor, N-methylcytisine and cytisine showed the highest affinity ofallquinolitidinealkaloidsstudied, confirming earlier work on the mode ofactionofcytisine(l,7). Lupanine, which is widely distributed in legumes as a
September 19941 Schmeller et af.: Quinolizidine Alkaloid Receptor Binding 1317 major alkaloid (2), displayed an IC,, of 5 p,M and is thus 100 times more active than hydroxylated lupanines or alkaloids of the multiflorine series. Quinolizidine alkaloids are also active at the muscarinic ACh-receptors, but often in a complementary fashion. Quinolizidine alkaloids with high affinities to the nicotinic receptor are less active at the muscarinic site and vice versa (Table 1). Especially active at the muscarinic receptor are 13a-tigloyloxylupanine, sparteine, angustifoline, albine, multiflorine, and 3P-hydroxylupanine. It has been established that cytisine acts as an agonist at the nicotinic receptor (1).Whether the other quinolizidine alkaloids have similar properties is not known at present since our assay only meaSured the affinityofalkaloidsfor the receptorsbut not agonistic or antagonistic activity.
A typical alkaloid mixture from lupines contains alkaloids that bind to both ACh-receptors. Because nicotinic and muscarinic receptors are widely distributed within the body, a number oftissues and organs will be affected, to a certain degree, ifananimal consumestheseplants. In addition, inhibition of Na’ or K+ channels, which has been found for sparteine and lupanine (1,7, M. Wink and R. Fink, unpublished data) might potentiate the toxicity caused by binding of the alkaloids to ACh-receptors. It has been proposed that most secondary metabolites do not have random structures, but that they have been shaped during evolution (“evolutionary molecular modeling”) to interact with cellular targets (2,7).Our study shows that quinolizidine alkaloids must have been optimized to more than one target since some quinolizi-
N-methylcytisine
0 sparteine
A anagyrine
0 10-~
10-~
io-*
io-’
ioD
io’
io2
io3
lo4
concentration of displacer IkM)
FIGURE 1.
Dose-response curves for the binding of some quinolizidine alkaloids to the nicotinic and muscarinic ACh receptorsas measured by the displacementof specifically bound radioligands. (Data represent means 2 standard deviation[n= 31). Figure 1A. Nicotinic receptor;ligand 3Hnicotine.
1318
Wol. 57, No. 9
Journal of Natural Products
100
90 80
8
70
z
2
50
x
40
4
30
a n
3 ;. &
20 10 0
lo-'
loo
10'
1o3
lo2
1o4
concentration of displacer [pMl
FIGURE1. Continued. Figure 1B. Muscarinic receptor; ligand 'H-QNB.
dine alkaloids can affect both nicotinic and muscarinic ACh-receptors whereas others bind Preferentially One Or the other.
EXPERIMENTAL TESTComuND%+umolizidinealkaloids were extracted from seeds of L.upinus albus, L. mutabilis, and Anagyris foeti& by previously de-
T ~ L 1.EBinding of Qunolizidine Alkaloids (IC5,,) to Muscarinic and Nicotinic Acetvlcholine Receptors. Compound
.............................. .................. Angustifoline ............................ Cytisine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3a-Hydroxylupanine ... 13~-Hydroxylupanine..................... Lupanine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lupinine . . . . . . . . . . . . . ... N-Methylcytisine . . . . . . . . . . . . . . . . . . . . . . . . .
......................
...................... ................ .... 13a-Tigloyloxylupanine . . . . . . . . . . . . . . . . . . .
Nicotinic receptor' 193 p M 2096 pM >500 p M 0.14 pM 190 pM 490 pM 5 CLM >500 pM 0.05 pM > 5 0 0 pM 155 FM 331 pM 310 pM 160 pM
Muscarinic receptor' 33 pM 132 pM 2 5 pM 400 pM 74 pM 140 pM 114 pM 190 pM 417 pM 47 pM 118 p M 21 pM 129 pM 11 pM
*IC,,, values indicate the concentration of a particular quinolizidine alkaloid that displaces 50% of the specifically bound radiolabeled ligand.
September 19941 Schmeller et al.: Quinolizidine Alkaloid Receptor Binding 1319 scribedmethods (2). The purificationand isolation of pure compounds was carried out by vlc using a CHCI,/CH,OH/”,OH gradient (11). When purified, the crude alkaloids (12 g) from L.albus afforded 6 g lupanine, 800 mg multiflorine, 300 mg 13a-hydroxylupanine, 200 mg 13a-tigloyland 13a-angeloyloxylupanine,150 mg albine, and 50 mg angustifoline. Tetrahydrorhombifoline and 3fbhydroxylupanine were isolated from L. mutabilis and cytisine, N-methylcytisine, and anagyrine from Anagyrisfwtilt. 17-Oxosparteine was prepared from sparteine (2), while sparteine and lupinine were purchased from Sigma. The identity of all quinolizidine alkaloids, which are analyzed in our laboratory routinely (2,12), was confirmed by capillary gc, gc-ms and ”C nmr.
was incubated for 40 min at 20” with differing concentrations of quinolizidine alkaloids or 1mM nicotine as a control. The GFlC filters were presoaked in BSA (1 mg/ml) to reduce non-specific binding of ’H-nicotine. Further procedures were the same as described above. ACKNOWDGMENTS We thank H. Staudter and H. Wurm for technical assistance, Mittex Anlagenbau (Ravensburg)for financial support, and K. Scheuer for helpful discussion.
LITERATURE CITED 1.
MEMBRANE PREPARATION FOR ACETYLCHOLINE RECEPTOR STLIDIEs.-Porcine brains, which
were obtained within 30 min after the death ofthe animals from a local slaughterhouse, were used to prepare receptor-rich membranes. The brains were immediately frozen in N,; 50 g brain per 200 ml ice-cold buffer (0.32 M sucrose, 10 mM K+phosphate buffer, p H 7.0; 1 mM EDTA) were homogenized twice for 15 sec in a blender and then for 1 minwithanultraturrax(l3).Thehomogenate was centrifuged three times for 15 min at 1400 g and 4” to separate cellular debris. The supernatant was spun down at 100,000 g for 60 min. The resulting pellet was resuspended in buffer (as above but without sucrose). Aliquots were stored frozen at -80’. Protein content was determined by the Biuret method, using bovine serum albumin as a standard.
2.
3.
4.
5. 6.
7.
BINDINGMAYs.-Binding assays (in triplicate) were performed using a rapid filtration technique essentially as described by Yamamura and Snyder (9).
8.
Muscarinirrec~tw.-Membrane preparations adjusted to 500 p g protein in a final volume of 500 p1 buffer were incubated with 3Hquinuclidinyl benzilate (QNB) (44.0Ci/mmol; Dupont NEN) for 1 h at 20” in the absence and presence of quinolizidine alkaloids, employing 2 pmol atropine as a blank. The incubation was stopped with 3 ml ice-cold 0.9% NaCl solution and filtered (by suction) through WhatmanGF/C glass fiber filters. The filters were washed three times with 3 ml 0.9% NaC1, placed in vials, and dried at 60”for 30 min. Their radioactivity was measured in a liquid scintillation counter (RackBeta, Pharmacia) using “Ultima Gold” (Packard) as scintillation cocktail.
9.
Nicotinirreceptw.-’H-Nicotine(64Ci/mmol; Dupont, NEN) was used to assay the specific binding of quinolizidine alkaloids to the nicotinic AChreceptor.The membranepreparation(asabove)
10.
11. 12.
13.
A.D. Kinghorn and M.F. Balandrin, in: “Alkaloids: Chemical and Biological Perspectives,”Ed.byS.W. Pelletier,John Wiley &Sons, New York, 1984, Vol. 2, pp. 105148. M. Wink, in: “Methods in Plant Biochemistry. Alkaloids and Sulphur Compounds,” Ed. by P. Waterman, Academic Press, London, 1993, Vol. 8, pp. 197-239. M. Wink,Them. AppI. Gen.,75,225 (1988). M. Wink, in: “Allelochemicals. Role in Agriculture and Forestry,” Ed. by G.R. Waller, ACS Symposium Series, Washington, DC,No. 330, 1987,pp. 524533. M. Wink, Planta M e d , 53, 509 (1987). M. Wink, in: “Insect-Plant Interactions,” Ed. by E.A. Bernays, CRC Press, Boca Raton, FL, 1992, Vol. IV, pp. 133-166. M. Wink, in: “The Alkaloids,” Ed. by G.A. Cordell, Academic Press, San Diego, 1993, Vol. 43, Chapter 1, pp. 1-1 18. M. Wink and T. Twardowski, in: “Allelopathy: Basic and Applied Aspects,” Ed. by S.J.H. Rizvi and V. Rizvi, Chapman & Hall, London, 1992, pp. 129-150. H.I. Yamamuraand S.H. SnyderJror. Natl. Acad. Sci., USA, 71, 1725 (1974). M. Wink, in: “Hagers Handbuch der PharmazeutischenPraxis,”Ed. by R. H-I, W. Keller, H. Rimpler, and G. Schneider, Springer-Verlag, Heidelberg, 1993, Vol. 5, pp. 623-630. S.W. Pelletier, H.P. Chokshi, and H.K. Desai,J. Nut. Prod., 49, 892 (1986). C. MeiBner and M. Wink, in: “Lupinen 1991-Forschung,Anbauund Verwertung,” Ed. by M. Wink, University of Heidelberg Press, Heidelberg, 1992, pp. 91-129. T. Haga and K. Haga, in: “Receptor Biochemistry. A Practical Approach,” Ed. by E.C. Hulme, IRL Press, Oxford, UK, 1990, Chapter 2, pp. 5 1-78.
Receiwd 1 7 March 1994
